SEELE AI / Resources
Ideas for building playable worlds.
Practical notes on AI games, creative workflows, and the tools that help ideas become something you can play.
How to Build a Survivors-Like Game with AI and Auto-Attacks
Build a short survivors-like run with automatic attacks, two waves, collectible XP, and one upgrade choice. Use a build prompt and exact test cases.
How to Make a Tactical Shooter with AI and Test an Encounter
Make a tactical shooter with AI. Build a two-position training encounter, test a teammate's timing cue, and check when the player can safely tag a target.
How to Build Your First Theme Park Management Game with AI
Build a small theme park management game with AI. Connect a path, run a carousel queue, track ticket revenue, and decide when to add a second ride.
How to Build a Twin-Stick Shooter with AI Using One Arena
Build a twin-stick shooter with AI by testing movement and aim separately in one arena. Use two simultaneous-input shots to verify direction and firing.
How to Build an Arcade Vehicle Combat Game with AI
Make a vehicle combat game with AI by building one arena around speed, braking, and a rear shot. Follow a step trace from approach to extraction.
How to Make a Word-Building Game with AI Using a Tile Rack
Make a word-building game with AI. Test a fixed letter rack, a short word list, scoring, and changing length rules across two turns.
How to Make an Inventory-Management Roguelike with AI
Learn to make an inventory-management roguelike with AI, focused on spatial item placement and build synergies. Plan the core loop, build a prototype and refine it.
How to Make a Monster-Taming Roguelike with AI
Build an original monster-taming roguelike with AI. Prototype creature capture, turn-based battles, team choices, and a boss encounter in one playable run.
AI Game Reverse Engineering: What Changes for Developers?
What AI can and cannot do in game reverse engineering, static recompilation, and ports—and what developers should verify about accuracy, rights, and originality.
JEV API Tutorial: From Game State to Real-Time Decisions
Learn how to design a JEV API integration with compact game state, legal actions, typed questions, response validation, deadlines, and deterministic fallback.
JEV Games: How JEV Participates in Real-Time Game Decisions
Explore JEV game demos and see where a typed decision model fits in the loop between game state, legal actions, execution, and fallback behavior.
How JEV Chooses Legal Actions in Game AI
Learn how to design a safe action space for JEV: generate legal actions in game code, let JEV choose or score them, then revalidate before execution.
JEV NPC: Real-Time Tactical Decisions for Game Characters
Learn how JEV can help an NPC choose between attack, retreat, healing, exploration, and assistance without replacing perception, behavior trees, or game execution code.
JEV Unity Tutorial: Build a Real-Time Tactical NPC
Build a Unity NPC architecture around JEV with a compact state, legal actions, a decision cadence, an executor, timeout handling, and an observable fallback.
JEV vs LLM: Which AI Should Handle Real-Time Game Decisions?
Compare JEV and LLMs for game AI. Learn when to use typed choices, scores, and yes-or-no judgments versus dialogue, planning, code, and narrative generation.
What Is JEV? A New AI for Real-Time Game Decisions
JEV is a bounded AI decision layer for games. Learn what JEV means, how it relates to game AI, and how it compares with LLMs, behavior trees, and finite-state machines.
JEV Game AI: Real-Time Tactical Decisions for NPCs
See how JEV fits into a game loop, helps NPCs choose between attack, retreat, healing, and exploration, and works alongside behavior trees and game code.
How JEV Helps Generate Games: From Game Ideas to Intelligent NPCs
Learn how JEV works with game-generation models, LLMs, code generation, and game engines to make generated NPCs more capable without replacing the engine.
JEV Tutorial: Build Your First Real-Time Decision-Making NPC
A practical JEV tutorial covering Game State, Legal Actions, decision calls, execution, cadence, latency, timeouts, and safe NPC fallbacks.
GPT-6 Astra 制作 AI Game 实践：从自然语言想法到可玩原型 | SEELE AI
用一套可验证的工作流，把自然语言游戏想法拆成核心循环、状态机和垂直切片，再交付可玩的 AI Game 原型。
UEFN MCP Setup and Toolsets Guide for 42.00 | SEELE AI
Set up UEFN MCP in Fortnite 42.00, choose supported toolsets, validate Verse and devices in PIC, and troubleshoot LUF/XYZ or init_unreal.py issues.
Game-Native AI Agents vs General Coding Agents | Seele AI
Game-Native AI Agents vs General Coding Agents. Use a focused workflow, observable evidence, and a clear human review boundary.
Comparing AI Game Platforms for Source Access and Commercial Release | Seele AI
Comparing AI Game Platforms for Source Access and Commercial Release. Use a focused workflow, observable evidence, and a clear human review boundary.
AI Coding Agents for Game Development: General Tools vs Game-Native Workflows | Seele AI
Compare general coding agents with game-native AI workflows for game development, including repository changes, playable prototypes, tests, handoffs, and production boundaries.